Taperssection.com

Gear / Technical Help => Post-Processing, Computer / Streaming / Internet Devices & Related Activity => Topic started by: airbladder on July 20, 2008, 10:59:53 PM

Title: Samplitude SE 9 workflow
Post by: airbladder on July 20, 2008, 10:59:53 PM
I recently purchased Samplitude SE 9 and am having some issues figuring it out.  Rite now I am stuck trying to resample from 48 to 44.1.  Every time I try it looks like nothing is happening.  After I click around a few times the screen turns white and on the top it says (not responding).  Any idea what I am doing wrong?
Title: Re: Samplitude SE 9 workflow
Post by: Brian Skalinder on July 20, 2008, 11:19:36 PM
After I click around a few times the screen turns white and on the top it says (not responding).

That doesn't sound good.  Though sometimes Windows confuses "busy" with "not responding" (typicaly a terminal condition for an app).

There are multiple ways to perform sample rate conversion, sometimes with differing results even though one uses the same quality settings.  This post lists the different ways to perform SRC:  http://taperssection.com/index.php/topic,99454.msg1324733.html#msg1324733

Which method did you use / what are your exact steps / what settings did you specify?
Title: Re: Samplitude SE 9 workflow
Post by: airbladder on July 20, 2008, 11:22:30 PM
Ok I think I got it.  I load my file, fade in, fade out, normalize, then export audio (wav).  Then go to format settings.  In format settings change the sample rate to 44.1 and set the format to 44.1 16 bit stereo.  Then I go into the dither (traing).  In there I leave the dithering with triangular spread nose (leave it at 0.50).  Then I go into resampling quality.  Under resampling quality I pick all of the highest settings.  Click Ok, ok, export.  Thanks Brian I will take a look at that link.
Title: Re: Samplitude SE 9 workflow
Post by: airbladder on July 20, 2008, 11:26:45 PM
I was using "<1>  Offline Effects | Change Sample Rate" when things freeze up.  I am doing this with a very small file (like 10 second).
Title: Re: Samplitude SE 9 workflow
Post by: Brian Skalinder on July 20, 2008, 11:48:02 PM
Odd that it seems to lock up when running offline.  Not sure what to say on that one.  You might search / tap the SAM forum to see if you can figure it out.  I'm at a loss.

In the meantime, you might also try the Track Bouncing method.  Unlike the Export method, it will allow you to specify the quality setting.
Title: Re: Samplitude SE 9 workflow
Post by: airbladder on July 21, 2008, 12:03:57 AM
When I go to track bouncing that option is not enabled.  I see it but it won't let me click on it.  I was able to get things to work by batch processing.  When I batch I can use ultra high 2.  But when you batch it does not give any dithering options.  Is is automatically using the highest quality dithering this way?
Title: Re: Samplitude SE 9 workflow
Post by: Brian Skalinder on July 21, 2008, 12:26:56 AM
Ahhhh...Track Bouncing is disabled because you must be loading the audio files directly into the WAV editing window (File | Load / Import).  I recommend doing all work in a Virtual Project (File | New Virtual Project).  Virtual projects (VIPs) allow you to apply non-linear, non-destructive, object-oriented edits - much better workflow than the linear edits available in the WAV editor.  If you need help working in a VIP, ask away - took me a while to ramp up the learning curve and get accustomed to it, but now I love it.
Title: Re: Samplitude SE 9 workflow
Post by: airbladder on July 21, 2008, 08:11:38 AM
So you recommend working in VIP and doing SRC through track bouncing.  Let me take a look.
Title: Re: Samplitude SE 9 workflow
Post by: Brian Skalinder on July 21, 2008, 10:14:40 AM
So you recommend working in VIP and doing SRC through track bouncing.  Let me take a look.

Working in VIP for all edits, including SRC, provides a much more powerful workflow.  SRC through Track Bouncing will let you specify the SRC quality.  I also recommend trying SRC from the VIP using the Offline method - see if the app locks up as it does from the regular WAV editor.

My high-level workflow is something like:


Alternatively, you could do similar, but use Offline SRC:

Title: Re: Samplitude SE 9 workflow
Post by: Gutbucket on July 21, 2008, 12:02:02 PM
^^^^^
Ahh just the starting point outline I needed.  [lights going on]

I've been sitting on Sam for a few months and just started taking a look at it over the weekend.  I was a bit confused after expecting the VIP type workflow but had opened the WAV file directly and found nothing but destructive editing happening. Thanks.
Title: Re: Samplitude SE 9 workflow
Post by: JoeKiller on July 21, 2008, 12:10:55 PM
Brian do you have any tips for normalizing sections of the source?  Like if we adjusted levels during the show.  I usually use a virtual project.
Title: Re: Samplitude SE 9 workflow
Post by: Brian Skalinder on July 21, 2008, 12:47:25 PM
Brian do you have any tips for normalizing sections of the source?  Like if we adjusted levels during the show.  I usually use a virtual project.

I've dealt with gain adjustments in two different scenarios:

<1>  Stepped gain adjustments

I've always found the point at which I made the stepped gain adjustment and split on a zero crossing into two objects  - one pre-adustment, one post-adustment.  I've then applied the appropriate fixed gain to whichever object requires it, typically the preceding object.

<2>  Variable gain adjustments

In this case, I've used the volume envelope to scale the gain adjustment by ear.  AFAIK, one may only apply the envelope to the entire track, so the envelope may rise up or down in an appropriate curve / line until the point at which I stabilized levels, and then the envelope remains flat for the remainder of the recording.

These both sound easy.  In the former scenario, it's not always easy to achieve a completely seamless adjustment, but I've found splitting objects on a zero crossing almost always worked fine.  In the latter scenario it's not always easy to immediately identify the points at which I start / stop the gain change, nor the line / curve required to make the adjustment.
Title: Re: Samplitude SE 9 workflow
Post by: boojum on July 21, 2008, 12:54:31 PM
^^^ What Brian says.  This is the way I work in SAM, too.  VIP is the way to go as if you screw something up it has not been applied.  It is applied when you do the track bounce.  I believe you have way more functions, too, in VIP.  If you have SAM 9 SE you are hampered by not having a hard copy of the manual for the software.  The on-line manual has the instructions for a fuller version, so that gets confusing.

SAM does have a long learning curve but all the editors do when they have a lot of functions.  They more they do the more you have to learn.  I like the editor.  You probably will, too, when you get familiar with it. 

Cheers

Title: Re: Samplitude SE 9 workflow
Post by: airbladder on July 21, 2008, 09:26:44 PM
So when it comes to dithering you leave it at "dithering with triangular spread noise, bit depth .50"?
Title: Re: Samplitude SE 9 workflow
Post by: live2496 on July 23, 2008, 09:10:08 AM
So when it comes to dithering you leave it at "dithering with triangular spread noise, bit depth .50"?

Yes. For the SE version that is what I would use. Samplitude support team recommends the setting of 0.5 for dithering. This is like 1/2 of a bit on average or something like that.

When working in the VIP, dither is automatically handled for you whenever the wordsize is reduced from float or 24-bits down to 16-bits. Like if you are bouncing to a 16-bit file or exporting to 16-bit format.








Title: Re: Samplitude SE 9 workflow
Post by: JAH on July 23, 2008, 09:28:42 AM
Absolutely....always go the vip route...this is what makes SAM so much better than Protools (IMHO).
Make changes, Then bounce, then resamp.
SAM is the bomb!


^^^ What Brian says.  This is the way I work in SAM, too.  VIP is the way to go as if you screw something up it has not been applied.  It is applied when you do the track bounce.  I believe you have way more functions, too, in VIP.  If you have SAM 9 SE you are hampered by not having a hard copy of the manual for the software.  The on-line manual has the instructions for a fuller version, so that gets confusing.

SAM does have a long learning curve but all the editors do when they have a lot of functions.  They more they do the more you have to learn.  I like the editor.  You probably will, too, when you get familiar with it. 

Cheers


Title: Re: Samplitude SE 9 workflow
Post by: airbladder on July 23, 2008, 11:43:45 AM
Can I get a work flow out line for mixing two 24/48 sources together that ran the same clock?  I feel like I am going through some unnecessary steps. 
Title: Re: Samplitude SE 9 workflow
Post by: Brian Skalinder on July 23, 2008, 12:40:04 PM
Can I get a work flow out line for mixing two 24/48 sources together that ran the same clock?  I feel like I am going through some unnecessary steps.

High level overview:

[01]  Create a new VIP at 24/48 with 2 tracks (assuming you have 2 stereo tracks).
[02]  Load Source 1 to Track1.
[03]  Load Source 2 to Track2.
[04]  Align Track1 and Track2.
[05]  Adjust relative mix of Sources 1 & 2 (multiple ways to do this - track volume, mixer, etc.).
[06]  Perform any other edits (fades, etc.)
[07]  Track.
[08]  Track Bounce (select target bit-depth and sample rate / apply SRC & dither;  results in single, mixed stereo file with track markers).
[09]  File | Export (this will simply cut up the file into tracks; specify same bit-depth and sample rate as the actual data so you don't SRC / dither again).
Title: Re: Samplitude SE 9 workflow
Post by: airbladder on July 24, 2008, 11:13:37 PM
When mixing two 24 bit sources together can you normalize both sources to 100% before you track bounce (dither, SRC, add edits, merge two source)?  Or do you need to normalize both files to -3db then track bounce combiningng the two sources will add up to being normalized to100%)?
Title: Re: Samplitude SE 9 workflow
Post by: Brian Skalinder on July 25, 2008, 01:19:13 PM
Can't normalize the source files first, since they're summed during track bouncing.  I always just selected the loudest peak, track bounced, noted how far off I was on the mixed level peak, then adjusted the source files accordingly.
Title: Re: Samplitude SE 9 workflow
Post by: live2496 on July 25, 2008, 10:31:11 PM
Can't normalize the source files first, since they're summed during track bouncing.  I always just selected the loudest peak, track bounced, noted how far off I was on the mixed level peak, then adjusted the source files accordingly.

Hi Brian,
I'm not sure if the SE version is the same with respect to this, but if you select "offline effects" and then "normalize" there is a radio button for "virtual" in that dialog. You can set the normalize level in db or percentage.

What this does is to simply calculate a new value for the volume in the object editor. The value is calculated when you press "OK".

Once two or more track/objects are set in this manner they could be bounced with closely matching volumes.

Of course, when doing this leave enough headroom for the mixed signal.

Gordon



Title: Re: Samplitude SE 9 workflow
Post by: airbladder on July 26, 2008, 12:08:17 AM
Yea, you "can" normalize each source before you track bounce.  I am wondering if you "should" normalize both sources independently before you track bounce.  Or if you wait until the two sources have been bounced/merged then normalize the final file.  I think what Brian is saying is not to normalize before you track bounce, normalize the final file.  But what I don't understand, having never mixed two sources together before, is how much does the volume increase when you combine the two sources.  Lets say you have two sources that both peak at -5db.  Then you mix them together.  It seems like the final source is going to be louder that -5db and you may have some overs.  How do you prevent that?  How much headroom do you need on each source before you track bounce so that you don't have overs in the final file?  That is what I was trying to figure out.  If I knew how how much the volume was going to increase when I track bounce I could normalize both 24 bit files accordingly before I track bounce.  For example, if the two sources were going to go up 6db when they are bounced I could normalize both of them to -6db before I  bounce.  Then after the files have been track bounced they would not need to be normalized because they would all ready be at 100%.   It seems like it would be better to normalize the 24 bit files rather than the final 16bit file.       

Thanks again for all your help and patience on this.  With out this forum I don't know how people like myself would get this information.
Title: Re: Samplitude SE 9 workflow
Post by: live2496 on July 26, 2008, 07:36:39 AM
If combining two sources you would need to leave about 6db of headroom on each.

You can preview in the mixer, so if you want to avoid levels from being to high you can audition it and watch the metering.

You don't need to do a 16-bit bounce until you have completed all adjustments. You can continue to work in 24-bit format or float format until then. You can also do bounces in float or 24-bit format at any time.

In Brian's workflow example, he is doing the adjustments in [5] and [6], so by step [8] it is ready for the bounce.
Title: Re: Samplitude SE 9 workflow
Post by: Gutbucket on July 28, 2008, 12:26:58 AM
..what I don't understand, having never mixed two sources together before, is how much does the volume increase when you combine the two sources.  Lets say you have two sources that both peak at -5db.  Then you mix them together.  It seems like the final source is going to be louder that -5db and you may have some overs.  How do you prevent that?  How much headroom do you need on each source before you track bounce so that you don't have overs in the final file?  That is what I was trying to figure out.  If I knew how how much the volume was going to increase when I track bounce I could normalize both 24 bit files accordingly before I track bounce.,,


The exact amount if increase depends on how closely correlated the material is.  If you mixed two identical copies of the same material together (completely correlated) the resulting sum would be 6dB louder.  If completely uncorrelated, the result would be about 1dB or so louder.  With what we're doing the material is likely more similar than not, so assuming a 6dB increase is safe.  How to best go about doing that and other general work flow issues with Samp is something I'm getting out of this thread since the program is also new to me. Is there a function that lists the highest and lowest sample values in the file and their positions in the file?

[edited spelling]
Title: Re: Samplitude SE 9 workflow
Post by: Brian Skalinder on July 28, 2008, 12:56:39 AM
Is there a function that lists the highest and lowest sample values in the file and their positions in the file?

There are at least couple quick ways to easily identify the peak sample value (though not its location):


There may be other ways to identify the peak value, and maybe even a way to locate its location, but I don't know 'em.  Dunno about the lowest.
Title: Re: Samplitude SE 9 workflow
Post by: live2496 on August 01, 2008, 11:30:37 AM
Also, when you bounce there is an option to show the maxiumum peak found in db.